Springtime Peaks of Trace Metals in Antarctic SnowMasaya Ikegawa,1 Mieko Kimura,I Kazuto Honda,2 Kazuo Makita,3 Yoshiyuki Fujil,4 Yoshinori Itokawa5IKyoto University Graduate School of Medicine, Department of Social Medicine, Environmental Medicine, Kyoto, Japan; 2Perkin ElmerJapan, Osaka, Japan; 3Takushoku University, Faculty of Engineering, Tokyo, Japan; 4National Institute of Polar Research, Tokyo, Japan;5Kyoto University Graduate School of Medicine, Department of Social Medicine, International Health, Kyoto, Japan

Drifting snow samples were collected at Asuka Sation (71°32'S, 24°08'E, 930 m above sea

level) over a period fiom July toD 1991; 36 elements (indudingN, M& K FCa, Fe, Al,

, V, Cr, Ml Ni, Cu, Zn, G,G S Sr, e, Pr, NL Sm, X td,

and SOIn

there was a proounced peak nc ion of most elements hewith non-.se saltslfate. Enriciment factr analys sgest that Na, Mg, Ca, and Sr arBe'o"arne orngan andAl, Fe, Mn, Rbi, Ni Ga, V. -and al the rare carthemnts are of or . Volcic

eption bhereaso

the precip Pb Cd, ad no-sea J inthespring at . y A tarctic snw,CeP-MS, NO

SO2-, trace metals, volcanic eruption. Environ Hea P t 105:654-59 (1997)

Because the polar regions are remote fromanthropogenic pollution sources, chemicalconstituents of Antarctic snow reflect thebackground level of the pollution in theatmosphere (1,2). Various substances aretransported over the Antarctic through theatmosphere (3) and deposited onto thesnow surface (4,5), but their mechanismsand possible sources are not yet fully under-stood. Particularly, polar stratospheric pre-cipitation in Antarctic snow has recentlyattracted many investigators to the connec-tion with the Antarctic ozone hole (6-8).

Chemical constituent concentrations inair and snow in the Arctic area show a dis-tinct seasonal pattern caused by a variationin long-range atmospheric circulation inhigh and midlatitude areas (9,10). Airbornesulfate and several trace metals of crustaland anthropogenic origin show strongpeaks in spring; this is known as Arctic haze(9,110). In the Antarctic, a seasonal patternof crustal, marine, and sulfate aerosol con-centration has been reported at the SouthPole (11-13) and at a coastal region (14)with a maximum for the crustal and sulfatespecies and with a minimum for the marinecomponent during austral summer. Butthere are few studies of snow in Antarcticathat clarify autumn-spring peaks of Pb inthe east coast of the Antarctic Peninsula(15) and summer peaks of NO3- and so42-concentrations at Mizuho Station (16).However, available data sets for long- andshort-term changes in chemical componentsin Antarctic snow are still limited.

Chemical constituent concentrations inair and snow in Arctic and Antarctic areasare extremely low. Previous analyses of traceelements in polar snow have been carried

out by graphite furnace atomic absorptionfollowing a preconcentration in a cleanroom (1,2,4,5,14). After the introductionof inductively coupled plasma-mass spec-trometry (ICP-MS), which is a well estab-lished powerful technique for the determi-nation of trace elements, it becomes possi-ble to determine as many as 40 elements atdetection limits of below the order of partsper trillion (ppt) (17,18). The present dataset, however, is probably one of the mostextensive so far reported for trace elementsin precipitation in the Antarctic. In thisstudy, we present temporal variations oftrace elements in the Antarctic area anddiscuss their using enrichment factors.

MethodsField sampling. From July to December1991, drifting snow samples were collectedat Asuka Station (71032'S, 24°08'E, 930 mabove sea level ). Snow was obtained by thesnow trap method reported by Osada at al.(16). Sampling was carried out with specialcare to avoid contamination. Asuka Stationis located 120 km from the coast in QueenMaud Land, East Antarctica (Fig. 1). TheS0r Rondane Mountains are situated sever-al hundred kilometers south of the station.At the station, katabatic winds prevail withthe mean wind speed of 12.6 m/sec in aneast southeast direction (19). Net accumu-lation of snow over January to November1991 at the station was estimated as 35 cmby the snow stake method (19); the devia-tion of the stake measurements was within10 cm in the period of this study (19.

Chemical analysis. Snow samples werecollected in specially prepared polyethylenebottles, which had been thoroughly

washed with nitric acid and distilled water.The samples were kept frozen until the ana-lytical stage. Sample pretreatment, such aspreconcentration or filtration, was not per-formed. Two sets of running conditions ofICP-MS (ELAN 6000, Perkin Elmer,Osaka, Japan) were used. For the analysesof Ca, Mg, K, and Al, the RF power was1000 W and the nebulizer gas flow rate was0.725 1/min; for all other elements, the RFpower was 600 W and the rebulizer gasflow rate was 1 1/min. Sampling cones andskimmer cones were made of Pt. Ultra purewater (Tamapure-100) and ultra-high puri-ty nitric acid prepared for the semiconduc-tor industry, both provided by TamaChemical Industry Co., Ltd (Tokyo, Japan)were used throughout the experiments.Dilutions were carried out on a clean benchwith teflon-coated volumetric ware. Weanalyzed the standard reference materialSRM1643c provided by the NationalInstitute of Standards and Technology forthe selected elements (17). Concentrationsobtained by this method were in goodagreement with the certified values for allthe elements'examined (17).

Besides major and trace metals, we alsomeasured C1, s042-, and NO3- concentra-tions in drift snow by ion chromatography(SHIMADZU HIC-6A, SHIMADZU,Japan).

Results and DiscussionSodium is often referred to as the marinereference element. The seasonal trend forNa shows a transient increase in earlyOctober, which is superimposed on a slowlyincreasing background from winter to sum-mer (Fig. 2). Previous meteorological stud-ies have shown that meridional long-rangeair transports from the surrounding oceansbecome prominent in September toOctober (20,21). Since a transient increasein wind speed was observed in the period of5-8 October in the present study (13), the

Address correspondence to M. Kimura, KyotoUniversity Graduate School of Medicine,Department of Social Medicine, EnvironmentalMedicine, Yoshida-konoe-cho 606-01, Sakyoku,Kyoto City, Kyoto, Japan.We thank all the members of 32nd JapaneseAntarctic Research Expedition, especially theWinter-over group at Asuka Station, for their helpand encouragement to our research.Received 16 October 1996; accepted 5 February 1997.

Volume 105, Number 6, June 1997 * Environmental Health Perspectives654

Articles * Trace metals in Antarctic snow

Na peak event can be attributed to sea saltspray transported during low pressure dis-turbances in the austral spring. The overalltrend in concentration of the chloride ionshows a good agreement with Na ([Cl] =1.71 x [Na]; r = 0.993). Chloride versussodium ratios for 18 samples were calculat-ed as 1.71, which is close to the sea watervalue of 1.73 (22).

The Mg, K, Ca, and Sr have peaks inthe spring that are significantly correlatedwith Na and Al (Tables 1 and 2). To exam-ine maritime contributions to these ele-ments, enrichment factors were calculatedby applying the following equation:

EFsea= [X/Na]snowl/[X/Na]sea

where X refers to the concentrations of theelements of interest (1,2,4,23). CalculatedEFsea for Mg, K, Ca, and Sr are almost inunity, except during late September whenthey are increased 2- to 10-fold. Therefore,origins other than sea salt spray must beconsidered during and after the peak eventfor these elements. Although a concentra-tion profile of Se has a significant correla-tion with Na, the contribution of the seasalt spray is almost negligible [Se/Na] =3.64 x 10-9) (22) and EFsea for Se is around102_104. From the studies of the composi-tion of volcanic smoke fumes, volcanoescould be a significant source of Se (24).

Aluminum, a crustal reference element,shows a strong peak in late September andearly October (Fig. 2). The second peak ofAl coincides with the Na peak. Except Naand Se, all of the elements examined have asignificant correlation with Al (Tables1-3). Crustal enrichment factors near unityare obtained for Fe, Mn, Rb, Cr, Ni, V,and all the rare earth elements before lateSeptember, suggesting that all these ele-ments are of crustal origin. High EFcrustvalues ranging from one to four orders ofmagnitude are observed for Pb, Cd, Cu,and Zn as previously reported (3). Thepossible natural source for these heavy met-als could be volcanism.

Lead concentrations for the recentAntarctic surface snow has been found tobe in the range of 2-13 pg/g (1,2,4,23,25).Irrespective of possible local lead emissionsfrom Asuka Station or snow vehicles, this isalmost the same value as that of drift snowfrom July to late September. The subse-quent lead peak may be due to volcanismor anthropogenic production transportedby long-range air transport processes.

Rare earth patterns are often used as trac-ers of atmospheric materials on a global scalebecause rare earth elements are chemically

Ice-free aa Moraine field Bem ice area

Figure 1. Asuka Station in Queen Maud Land, East Antarctica

2.5

Eg 2Q

a.*CoCD' 1.5zC.,

z

li I.L

ov 0.5Q

July Aug Sept Oct Nov Dec

1991

Figure 2. Seasonal concentrations of Na, Al, Cl-, NO3-; and So42- in snow at Asuka Station.

Environmental Health Perspectives * Volume 105, Number 6, June 1997

250

200

150 .=1

a.

100 S0a0

50

655

Articles - Ikegawa et al.

Table 1. Seasonal variations in the concentrations of major elements

Date Na Mg K Ca Fe Al

25 Jul-20 Aug 3.25 x 102 2.58 x 10 2.55 x 10 1.52 x 10 1.49 x 10 6.0820-25 Aug 8.90 x 10 9.83 9.28 7.68 5.84 x 10-1 4.6225-30 Aug 1.45 x 102 1.78 x 10 1.13 x 10 8.71 7.26 x 101 4.6630 Aug-5 Sept 1.50 x 102 1.87 x 10 1.18 x 10 1.04 x 10 1.18 4.825-10 Sept 2.11 x 102 2.42 x 10 1.58 x 10 1.54 x 10 3.40 5.3410-24 Sept 3.01 x 102 3.13 x 10 2.81 x 10 3.45 x 10 3.81 6.9224-27 Sept 4.38 x 102 1.45 x 102 1.90 x 102 3.52 x 102 1.63 x 102 8.53 x 1027 Sept-3 Oct 3.73 x 102 9.09 x 10 7.89 x 10 3.05 x 102 3.53 x 10 2.74 x 103-8 Oct 1.13 x 103 2.19 x 102 3.94 x 102 6.32 x 102 1.60 x 102 8.36 x 108-16 Oct 4.38 x 102 6.01 x 10 7.28 x 10 1.23 x 102 3.51 x 10 1.50 x 1016-24 Oct 6.50 x 102 1.06 x 102 7.51 x 10 9.88 x 10 4.35 x 10 2.08 x 1024 Oct-2 Nov 8.72 x 102 1.25 x 102 8.72 x 10 1.68 x 102 3.51 x 10 1.87 x 102-8 Nov 9.84 x 102 1.14 x 102 1.06 x 102 2.28 x 102 7.58 8.638-14 Nov 7.99 x 102 9.99 x 10 8.44 x 10 1.90 x 102 1.37 x 10 1.27 x 1014-21 Nov 1.30 x 103 1.81 x 102 1.08 x 102 1.77 x 102 2.23 x 10 1.92 x 1021-28 Nov 9.32 x 102 1.41 x 102 1.01 x 102 2.39 x 102 2.02 x 10 1.70 x 1028Nov-7Dec 6.16x102 9.14x10 7.91 x10 1.93x102 1.84x10 1.55x107-22 Dec 1.43x 102 2.88 x 10 2.74 x 10 6.39 x 10 1.75 x 10 1.58 x 10Detection limit 0.03 0.007 0.015 0.05 0.005 0.006Naa 1.0000 0.8989** 0.6397** 0.6390** 0.3189 0.3383Ala 0.3383 0.6955** 0.8747** 0.8465** 0.9909** 1.0000

Detection limits (DL) are defined here as the equivalent concentration of three times the standard devia-tion of the blank response (unit = 10-9 g/g).aCorrelation coefficient with Na and Al by paired t-test.**p<0.01.

similar to each other and arise from the samesources (26,2X. In the present study, a gen-eral trend of rare earth elements is character-ized by bimodal peaks in late September andearly October, as is Al (Table 3). Before peakevents, the rare earth pattern in snow is simi-lar to that in sedimentary rock; thereafter,this pattern becomes more like the rare earthpattern in acidic rock, according to Taylor'stable of elment abundance (28). A crustalenrichment factor for all the rare earth ele-ments was <5 before the peak events, and theEF ctc for La, Ce, Pr, Nd, Sm, Gd, and Th

increased to 5-10 after the peak events.Local emissions from the Sor RondaneMountains and long-range air transportprocesses could be the source of rare earthelements in the drift snow.

It is customary to refer to the SO42- forpolar snows as non-sea salt (nss) s042,where nss S042- is the non-sea salt concen-tration corrected for the marine contribu-tion (29b. Apparently, nss 5042- shows verylow or even negative concentration until24-27 September; thereafter, it increases ashigh as 150-650 ppb (Fig. 3). Previous

Z0-

0CO

July Aug Sept Oct Nov Dec

1991

Figure 3. Seasonal concentrations of non-sea saltSQ42- in snow at Asuka Station.

studies have shown that volcanogenic so42-is the dominant component following majoreruptions in Antarctica (30). Another possi-ble source for nss sO42- is the biogenic pro-duction in the oceans. Non-sea salt sulfate,generally in the form of sulfuric acid orammonium sulfate, is a precursor that couldbe from dimethylsulfide released by the bio-genic activity of the marine surfaces or car-bonyl sulfide, as suggested by others (31). Asignificant correlation of nss so42- with Naor Cl- in this study supports this hypothesis.

From July to late September, NO3- con-centration stayed below 60 ppb, graduallyincreased to 570 ppb in mid-November,and recovered in December. Spring to earlysummer maximum levels in nitrate have alsobeen confirmed in the South Pole ice cores(6-8). The sublimation process could notexplain this trend very well because a maxi-mum sublimation was thought to occur inmid- to late December, when the daily solarradiation showed its maximum (32). NO3-versus nss SO42- concentration shows posi-

Table 2. Seasonal variations in the concentrations of trace elements

Date Li V Cr Mn. Co Ni Cu25 Ju-20 Aug 1.34 x 10 1.31 x 10 <DL 6.24 x 10 2.88 <DL 4.16 x 10

gt--luBYt .10.....!137 - a il w i is lJBx10.....................25-30 Aug 6.15 1.31 x 10 <DL 3.15 x 10 1.21 <DL 3.59 x 10a Alt'j'lIEIW;0; -lill§!gAs sj4; 3i1 ;"111! ii ii :i li 11 "Ii11 -3.S~~QM; IN 3N..O-10 Sept 8.79 1..20 x 10 <DL 2.92x10Z 1.91 <DL 3.94x 10

4.02 x 162 7.06 x 102 4.73x 102 5.46 x 1io 1.64Oxt1 3.55x 102 2.84x 102

3-8Oct 5.92x16 807x1 4.33x1 6. 10 1.70x102 4 1x12 2.88X1024. - ,"IW It "W'x1s:7.57x1

M~~~~~~~~~~~~~~47 Ix Uf '

lli:Jrg#sisMsNOwXsFRUswiEtsi-rS !iiltGs~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~53.06 x 102 .19x11$ 1.98x 1 3 3.87x10 1..14x 102 1.26x 102

,ggjlrg.p, W, qM. t'll .,k lip.gj p-I'm; 4 -.Nil- N:: "f!"............m

2-8 Nov I x 6.53 x 10 3.59 x 10 1.5 4.49 x 10;f d

..... .....

14-21 Nov 1.45 x I 1.14 x liP 2.49 x 10 6.07 x 10::i,,m-wn_g gmg N...

10228 Nov-7 Dec 1.04 x 102 5.37 x 10w:: i4l f, W

i- ... .. ......HM I:s,!:ff'!.:f:::::.l.::..::::..;:us;:,.

ii!;::If;.."fs...:sf.l:::.-.:.,-:.A --t-Ms fi! g AWkwiM." .%.;;.::: :: ::t:::UtfM ::f,.;:f.ft; ':;:f-t:::.,.:::.- :.. "..1%'IMERNMEM. .. .. .... .... gmppgg' .. ... .....N

......... .1 MW .....

Detecfion imit 0.002 0.02 0.002 0.0009.. ni..pw :;:..Rji;lW B tyfiy.. P w ff;'rw "4

q'TM PO'N TR .0011'r iTt SN.s f::: ii;. f: :f :M f:.;MIUM. .0S, S-;, t;M: Ztt TM M Pt

r

Al" 0.93 0.9608** 0.9929** 0.9288** 0.9614**

Volume 105, Number 6, June 1997 * Environmental Health Perspectives

Detection limits (DL) are defined here as the equivalent concentration of three times the standard deviation of the blank response ( unit = 109 g/g).aCorrelation coefficient with Na and Al by paired t-test.*p<0.05; **p<0.01.

.:

24-27 Sept

16-24 Oct 1........... .....

656

Articles - Trace metals in Antarctic snow

tive correlation in the present study. Recentwork indicates that nitrate can enter the tro-posphere through stratosphere-troposphereair exchange or through the subsidence ofnitrate-laden polar stratospheric clouds. Theincrease of tropospheric HNO3 could bereflected in spring polar precipitation duringor after October, and the intensity of thisprocess may depend on the temperature ofthe polar vortex (6-8).

From the meteorological data collectedat Syowa Station in 1991 (19), the lowestvalue of daily total ozone was observed on30 September. This ozone loss in the lowerstratosphere is believed to be a new phe-nomenon peculiar to the 1991 polar vor-tex, and its relationship to volcanic activityhas been suggested by Hofmann et al. (33).The eruption of Mt. Pinatubo (15N,120E, 13-15 June 1991) caused what isbelieved to be the largest aerosol perturba-tion to the stratosphere this century (34).In addition, in the austral spring of 1991,the Antarctic lower stratosphere was char-acterized by a layer of volcanic aerosolsfrom the Cerro Hudson eruption (46S,73W, 12-15, August 1991) (35).

Lower tropospheric ozone followed aprominent seasonal change of winter maxi-mum and summer minimum, which isvery close to results of surface ozone mea-surements (36-38). It has been suggestedthat the intrusion of stratospheric ozoneinto the troposphere occurs over SyowaStation throughout the year and that airwith low ozone concentration is transport-ed from subpolar or middle latitudes toAntarctica through the lower tropospherefrom spring to early autumn.

In summary, a bulk deposition of mostof the elements and non-sea salt sulfatecould be transported by a long range airtransport process or polar stratospheric pre-cipitation in austral spring at AsukaStation. Particularly, it reflects the atmos-pheric peculiarity of austral spring in 1991,characterized by volcanic emissions of Mt.Hudson and Mt. Pinatubo. Because onlyone season of drifting snow data is availablein this study, it should be emphasized thatthis phenomenon must be confirmed byfurther studies and snow pit chemistry.

REFERENCES

1. Wolff EW, Peel DA. The record of global pol-lution in polar snow and ice. Nature313:535-540 (1985).

2. Landy MP, Peel DA. Short-term fluctuations inheavy metal concentrations in Antarctic snow.Nature 291:144-146 (1981).

3. Lambert G, Ardouin B, Sanak J. Atmospherictransport of trace elements toward Antarctica.Tellus 42B:76-82 (1990).

4. Weiss VW, Herron MH, Langway CC Jr.Natural enrichment of elements in snow.Nature 274:352-353 (1978).

5. Boutron C, Lorius C. Trace metals in Antarcticsnow since 1914. Nature 277:551-554 (1979).

6. Mayewski PA, Legrand MR. Recent increase innitrate concentration of Antarctic snow. Nature346:258-260 (1990).

7. Legrand MR, Kirchner S. Origins and varia-tions of nitrate in South Polar precipitation. JGeophys Res 95:3493-3507 (1990).

8. Mulvaney R, Wolff EW. Evidence forwinter/spring denitrification of the stratospherein the nitrate record of Antarctic firncores. JGeophys Res 98 D3:5213-5220 (1993).

9. Davidson CI, Jaffrezo JL, Mosher BW, DibbJE, Borys RD, Bodhaine BA, Rasmussen RA,Boutron CF, Ducroz FM, Cachier M. Chemical

constituents in the air and snow at DYE 3,Greenland. Atmos Environ 27A 17/18:2723-2737 (1993).

10. Heidem NZ, Wahlin P, Kemp K. Arcticaerosols in Greenland. Atmos Environ 27A17/18:3029-3036 (1993).

11. Bodhaine BA, Deluisi JJ, Harris JM. Aerosolmeasurements at the South Pole. Tellus38B:223-235 (1986).

12. Cunningham WC, Zoller WH. The chemicalcomposition of remote area aerosols. J AerosolSci 12(4):367-384 (1981).

13. Sukegawa Y, Yamanouchi T. Meteorologicaldata at Asuka Station, Antarctica in 1991. JARERep 190 (Meteorology 30) (1993).

14. Wagenbach D, Gorlach V, Morser K, MunnichKD. Coastal Antarctic aerosol: the seasonal pat-tern of its chemical composition and radionu-clide content. Tellus 40B:426-436 (1988).

15. Suttie ED, Wolff EW. Seasonal input of heavymetals to Antarctic snow. Tellus 44B:351-357(1992).

16. Osada K, Ohmae H, Nishio F, Higuchi K,Kanamori S. Chemical composition of snowdrift on Mizuho Plateau. Proc NIPR SympPolar Meteorol Glaciol 2:70-78 (1989).

17. National Institute of Standards and Technology.Standard Reference Material 1643c.Gaithersburg, MD:National Institute ofStandards and Technology, 1991.

18. Yamasaki S, Tsumura A, Takaku Y. Ultratraceelements in terrestrial water as determined byhigh-resolution ICP-MS. Microchem J49:305-318 (1994).

19. Abe T, Iwamoto M, Sukegawa Y, Inayoshi H,Aono M. Meteorological observations at SyowaStation and Asuka Station in 1991 by the 32ndJapanese Antarctic Research Expedition. AntarctRec 38(3):268-321 (1994).

20. Fujii Y, Ohata T. Possible causes of the varia-tion in microparticles concentration in an icecore from Mizuho Station, Antarctica. AnnGlaciol 3:107-112 (1982).

21. Ito T. Study on properties and origins of aerosolparticles in the Antarctic atmosphere. Papers in

Zn Ga Se Rb Sr Cd Pb4.85 x 10 5.19 <DL 2.09 x 10 2.28 x 102 <DL 1.28 x 10,6.3x , .44-.,'D,'; 8'.3 DL -. --:5.07x 10 2.10 <DL 1.40 x 10 1.24x 102 DL 1.22x10

..... ..;t..w:t~l7 "'': :......" ";i 5.17 41 1.43x10 : I1.24x92<DL 1.22xlD04.61>x0 4.81 DL 1.40xl0 1.78x102 cDL 1.05x10

1.6x103 2.96x102 <DL 1.32x103 2.10x103 4.34 1.35x102

9.19x102 3.40>x102 1.30x102 1.61 x103 3.38x103 4.55 9.68>x10....tx... LUs *. . L : .. 1x 2 - ~ O<X:.... .. ;.05.38x 102 9.47x10 1.32x 102 2.18 xz10 8.73x 102 <DL 4.37x10

1.29x102 5.71x t 1.33x 102 1.5x102 1.48>x10 cDL 1.22x 10

:~~~~~~10x. I........xWl : 2Q w

3,24>x102 9.16>10l 1.71>x102 2.33>x102 2.0910x<o DL 3.60>x10.. ~ Zh x-..xt--.-40 --->10 -.1x -~t...:.:;28--::.:x.l;O

2.06x102 1.09x 10 1.26x102 2.48x102 1.19x 103 cDL 2.86x 10

~~~~~~~~.1..1.5403 25>~3.0x102 <31 573103.x43 DL4

0.003 0.001 0.06 0.003 0.0008 0.003 0.001.i .i .; ..?Q.9534** 0.9799** 0.0066 0.9849** 0.8038** ossi6i11 0.6395**

Environmental Health Perspectives * Volume 105, Number 6, June 1997 657

Articles * Ikegawa et al.

Table 3. Seasonal variations in the concentrations of rare earth elements

V La Ce Pr Nd Sm Eu G

25Jul-2OAug 1.53 6.75 1.21x10 1.16 3.93 <DL <D DI~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~s._-_~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~......

25-30OAug 1.20 2.47 5.62 5.51x101 2.18 <DL <D DJul.MU, API.,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.......5-lO~~~~~~~~~~~~~~~~~~~~~ept1482.985.03694>10.12.39 <DL <DL <DL~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~......

4if~~~~~t~~ifi~~~~Thr~~~~iI~~~~taif~~~~.~~~-4iiifZ4~~~~~~~~ - -I", Seaif i.Ai

~~~O*~~~~Piw tl~~~~~~~~~~l~~~tI 1t01. 2 ""U*~~~~~~~~~~~~~~~~~~~~~W it 910KW ISO!Ig2y!=: i~

M~~~~~~~~~- A=~SrF ?M 1 ~ IDet0 etio lii1.000 0.098 0.000 <0.00051O 0.002 0.00 0010.2

q~p~t..ifivniIII 22 ~ j4I~ p~E 4.viI"LtlEI I Ijl -~~~~~~~~~ "I ~~~v A10.8299""o~~~~~~~~~~~~~~~eeez""0.8341"" 0.841?" 0.6810""0.7290"" 0.8355"" 0.7573""~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.!;Detection limits(DL)aredefined here as the equivalent concentration of three times the standarddeviationofthe~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Rd, t

blankresponse(unit-iO~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~g/g).~~~~~~~~~~~~~~~~~~!5a~~~orrelationcoefficient with Na and Al by paired t-test.~~~~~~~~~~~~~~~ ... - N-E~~n

""p<O.01.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~2 :

Meteorol Geophysics 34(3):151-219 (1983).22. Duce RA, Arimoto R, Ray BJ, Unni CK, Harder

PJ. Atmospheric trace elements at EnewetakAtoll. J Geophys Res 88:5321-5342 (1983).

23. Peel DA. Trace metals and organic compounds inice cores. In: The Environmental Record inGlaciers and Ice Sheet (Oeschger H, Langway CCJr, eds). Chichester, UKWiley, 1989;207-223.

24. Mroz EJ, Zoller WH. Composition of atmos-pheric particulate matter from the eruption ofHeimaey, Iceland. Science 190:461-465 (1975).

25. Boutron C, Patterson CC. Relative levels of nat-ural and anthropogenic lead in recent Antarcticsnow. J Geophys Res 92:8454-8464 (1987).

26. Tuncel G, Aras NK, Zoller WH. Temporalvariations and sources of elements in the SouthPole atmosphere. J Geophys Res 94 D1O:13025-13038 (1989).

27. Olmez I, Gordon GE. Rare earths: atmosphericsignatures for oil-fired power plants and refiner-ies. Science 229:966-968 (1985).

28. Taylor SR Abundance of chemical elements in

the continental crust: a new table. Geochimicaet Cosmochimica Acta 28:1273-1285 (1964).

29. Clausen HB, Langway CC Jr. The ionicdeposits in Polar ice cores. In: TheEnvironmental Record in Glaciers and Ice Sheet(Oeschger H, Langway CC Jr, eds). Chichester,UK:Wiley, 1989;225-247.

30. Legrand M, Delmas RJ. A 220-year continuousrecord of volcanic H2SO4 in the Antarctic icesheet. Nature 327:671-676 (1987).

31. Delmas RJ, Briat M, Legrand M. Chemistry of

.III Excellence in basic research'# at the National Institute of Environmental Helth Sciences

t4JeIDIG.; National ToxicologyProg ram

The National Toxicology Program, headquartered at NIEHS, evaluates emirMal age aspossible causes of cancer, birth defects, or genetic mutatons...rswone-third of such tests conducted anywhere in the world...adnd see to pregulatory agencies base their risk assessments and public health re a b Tare leading a world-wide effort to develop quicker cheaper and lessaniclassic 2-year rodent bioassays.

A part of the National Institutes of Health, the National Institute of Environmentia Health Sciences is loeinResr Trial

658 Volume 105, Number 6, June 1997 * Environmental Health Perspectives

Articles * Trace metals in antarctic snow

Tb Dy Ho Er Tm Yb Lu<DL <DL <DL <DL <DL <YL <DL

A Oak! it, -f O AWM- 'inhibited.......I...

<DL <DL <DL <DL <DL <DL <DLum. s,: i-;l uX,m

1,Nz W.,MWIMl10011.II Qlkl`:-'ll at O'.UL

<DL <DL <DL <DL <DL <-b-L <..DL

.................N11

E spB. R", abortiono u ;f! ILM"10 .N - :!.... = :f M, M p:tmgmi-tig 'grinmn kN. 10 W. M M-M tt M E P

.... ..... ... .... ......2.26 9.61 1.53 3.82 5.44 x I O' 3.80 4.25 x I 0-l'air

.. I-

M5 ..20in,

MP.WMIMt -,Scott, Axf.

"Tt I-JUW,'t 'M f,

1.32 5.96 9.43 x 1 0-1 1.63 2.27 3.13 x I O'Circulation. Stan,

::.:4"::.:. ',f -Y..

17, a "IU... .... .;.fXM i, ...... ..... .. .. .........

mas IW_.

6",'9' X_ 10- 4.07 k4-7 x 10- 5.30 <DL 1.28 2.43 x ITFij i:

'f MpINM EM

<DL <DL <DL <DL -<DL <DL <DLe,!.-;.xPennsylv ania' -.:,...:,-f;l:!l::::..-,l",.l -.l.."f..Ml-,: ---. .. l..... --.. :; .:::::::;:f:::. .. :::'.::.:, i:fMartinique Alm

. ..... .......... .. .... .... ..lwm,aIll"Irl at IN O"t6.93 x 1 O' la tt.,70 x lff .......... ..... <DL- <-D" L""' 1.41 1.43 x 10-1 1.39 x 10

igggimigg gggg :f:. ;;:.Mf.:.'.. -.::::.::,..:- ..%MR t:

ff:::..-'. ;. ln:s:file;WS IffiRm L iM M

3. x IC- 1.97 3.66 x <DL <DL <DL <DL"ag'. tl .: ii. ;:. ..:: W -iEl"! .l Q4 R ".4x Vgl'i rf t4lg1g. Italian'.,, .Mw

MMit Willi,

.......... ....... .. .......... .............. ... .. ... ...... ... .. ..... .. ......... ... ... ... ... ..... .timi n g; :f::,:. .... .... ..... ... ......... ....... ....... .... ..... .... ....... .. .i" .. ..........<0.0005 0.001 <0.0005 0.0008 40005 0.001 4=5

s::;t;:- f...t ... s;Vss:

tsi uh VM.tM :, t`5Mij n:,;I:fl; tO......... .... Li

..0 f.

0.7932** 0.8313** 0.571 6* 0.606 0.8077** 0.8291**

South Polar snow. J Geophys Res 87:4314-4318 (1982).

32. Fujii Y, Kusunoki K. The role of sublimationand concentration in the formation of ice sheetsurface at Mizuho Station, Antarctica. JGeophys Res 87:4293-4300 (1982).

33. Hofmann DJ, Oltmans SJ, Harris JM, SolomonS, Deshler T, Johnson BJ. Observation and pos-sible causes ofnew ozone depletion in Antarcticain 1991. Nature 359:283-287 (1992).

34. McCormick MP, Thomason LW, Trepte CR.

Atmospheric effects of the Mt. Pinatubo erup-tion. Nature 373:399-404 (1995).

35. Scott DD, Gregg JSB, Schnetzler CC, KruegerAJ, Walter LS. Circumpolar transport of theCerro Hudson SO2 clouds. EOS 72, 45,5:489-504 (1991).

36. Murayama S, Nakazawa T, Tanaka M, Aoki S,Kawaguchi S. Variations of tropospheric ozoneconcentration over Syowa Station, Antarctica.Tellus 44B:262-272 (1992).

37. Schnell RC, Liu SC, Oltmans SJ, Stone RS,

Hoffman DJ, Dutton EG, Deshler T, SturgesWT, Harder JW, Sewell SD. Decrease of sum-mer tropospheric ozone concentrations inAntarctica. Nature 351:726-729 (1991).

38. Aoki S. Syowa-kichi niokeru tairyu-ken ozone-noudohenka. Kyokuchi 57:1-4 (1993).

International Conference

Volatile Organic Compoundsin the Environment, _

Risk Assessment and Neurotoxicity

Pavia, ItalyOctober 5-7, 1997

Organized by the University ofPavia and itthe International Society ofthe Built Environment

A _Tej43

Environmental Health Perspectives * Volume 105, Number 6, June 1997 659